Observable island identification for state estimation using incidence matrix (Q1094366)

This paper deals with the topological observability analysis and the development of an observable island identification algorithm in large power systems by using the incidence matrix. The topological observability is analyzed by using the partitioned incidence matrix which has all the information on the system topology and measuring points as well. The topological observability test is performed by applying the integer Gauss elimination method to the topological measurement equation. For the identification of observable islands, a bus grouping algorithm is developed on the basis of the topological observability analysis under the assumption of all measurements being zero. This algorithm remarkably reduces the computational burden compared with the conventional algorithms. A meter placement algorithm is presented to recover the observability of an unobservable measurement system.